Kuningas, M.
Citation
Kuningas, M. (2007, December 4). A study into genes encoding longevity in humans. Retrieved from https://hdl.handle.net/1887/12474
Version: Corrected Publisher’s Version
License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden
Downloaded from: https://hdl.handle.net/1887/12474
Note: To cite this publication please use the final published version (if applicable).
CHAPTER 3
The Liver X Receptor Alpha Associates with Human Lifespan
Simon P. Mooijaart, Maris Kuningas, Rudi G. J. Westendorp, Jeanine J. Houwing- Duistermaat, P. Eline Slagboom, Patrick C. N. Rensen and Diana van Heemst The Journals of Gerontolog y Series A: Biological Sciences and Medical Sciences (2007) 62A, 343–349
Abstract
In the nematode Caenorhabditis elegans, nuclear hormone receptor DAF-12 regulates the deci- sion to go into a resistant dauer diapause, in which the worm exhibits a decreased rate of aging.
Using sequence similarity searches, we previously identified the liver X receptor alpha (LXRA) as one of the human nuclear hormone receptors the protein sequence of which is most similar to C. elegans DAF-12. Here, we studied whether variation in the gene encoding LXRA associates with human life span. In the Leiden 85-plus Study, a population-based prospective follow-up study, we genotyped four polymorphisms spanning the gene coding for LXRA (NR1H3) and tagged four common haplotypes. Among 563 participants, haplotype 2 associated with reduced mortality during the 7-year follow-up (hazard ratio 0.78; p=0.015), predominantly caused by re- duced mortality from infectious disease (hazard ratio 0.31; p=0.023). Haplotype 2 also associated with higher levels of plasma apolipoprotein E, a target gene of the LXRA (p=0.018), and higher levels of triglycerides (p=0.041). Genetic variation in the gene coding for the LXRA (NR1H3) associates with human lifespan.
Introduction
Human lifespan is under genetic control (Schoenmaker et al., 2006; vB et al., 2006), but only few specific genes modulating lifespan have been identified. In the nematode worm Caenorhab- ditis elegans, DAF-12 is a nuclear hormone receptor (NHR) that in response to environmental cues regulates the entry into dauer diapause (Antebi et al., 1998). Under adverse environmental conditions, unliganded DAF-12 coordinately turns down essential traits—such as metabolism, feeding, and reproduction— making the worm more stress resistant and extending larval sur- vival up to 5-fold (Klass and Hirsh, 1976), which suggests that during the diapause the worm ages at a lower rate. Genetic mutations in daf-12 can be either dauer defective or dauer constitu- tive (Antebi et al., 2000) and, in parallel, can decrease or increase adult life span of C. elegans (Fisher and Lithgow, 2006). Using sequence similarity searches, we previously identified the liver X receptor alpha (LXRA) as one of the human NHRs the protein sequence of which is most similar to C. elegans DAF-12 (Mooijaart et al., 2005). However, nothing is known about the as- sociation of genetic variants in the gene coding for the LXRA (NR1H3) with human lifespan.
In humans, the LXRA is expressed in the liver, kidney, macrophages, astrocytes, and other tissues (Peet et al., 1998). Oxysterols are breakdown products of cholesterol and serve as ligands for the LXRA (Janowski et al., 1999). Binding of ligands leads to the transcription of target genes that coordinately regulate various processes that together result in increased catabolism and excretion of cholesterol from the body (Lu et al., 2001; Repa et al., 2000). In humans, cholesterol is a major determinant of mortality in old age, especially from infectious disease (Weverling-Rijnsburger et al., 1997).The LXRA is also involved in innate immunity, as activa- tion of human macrophages that produce cytokines is dependent on LXRA (Joseph et al., 2003).
In humans, cytokine production is a highly heritable characteristic (de Craen et al., 2005) and associates with diseases and mortality up to the highest age category (van den Biggelaar et al., 2004a). These observations make the LXRA a candidate to affect human life span.
To test the hypothesis that the LXRA is involved in modulating human lifespan, we made use of genetic variation in the gene coding for the LXRA (NR1H3). Out of the data that recently came available from the HapMap Project, we selected four evenly spaced haplotype-tagging sin- gle nucleotide polymorphisms (SNPs) spanning the NR1H3 gene. In the Leiden 85-plus Study, a prospective population-based follow-up study of 563 elderly persons aged 85 years or older and onwards, we studied the association of the common haplotypes with survival during a mean follow-up period of almost 5 years. To further explore a potential role of the LXRA in biological mechanisms associated with modulation of human lifespan, we associated the genetic variation to mortality-related phenotypic markers.
Participants and methods
Participants
The Leiden 85-Plus Study is a prospective population-based study, in which inhabitants of Leiden, The Netherlands, aged 85 years, were invited to take part. There were no selection
criteria related to health or demographic characteristics. The study population consists of 599 individuals (all members of the 1912–1914 birth cohort) who were enrolled in the month of their 85th birthday between 1997 and 1999 (der Wiel et al., 2002). DNA was available for 563 people.
The Medical Ethical Committee of the Leiden University Medical Center approved the study, and written informed consent was obtained from all participants.
Causes of death
All participants in the Leiden 85-plus Study were followed for mortality until August 1, 2005.
Primary causes of death were obtained from death certificates registered at the Dutch Central Bureau of Statistics and categorized according to the 10th International Classification of Dis- eases (ICD). Specific causes of death were categorized into cardiovascular disease (ICD codes I00–I99), infectious disease (ICD codes A00–B99 or J11–J18), cancer (ICD codes C00–D48), or other causes (all other ICD codes).
Plasma measurements
At baseline, participants were visited twice at their place of residence within 1 month after their 85th birthday. All blood samples were collected early in the morning, but fasting was not required.
Plasma levels of total cholesterol, high-density lipoprotein (HDL) cholesterol, triglycerides, and C-reactive protein (CRP) were analyzed on fully automated computerized analyzers (Hitachi 747 and 911; Hitachi, Ltd, Tokyo, Japan). The level of low-density lipoprotein (LDL) cholesterol was estimated by the Friedewald equation (LDL cholesterol [mmol/L] ¼ total cholesterol – HDL cholesterol – [triglycerides/2.2]), whereby participants with a triglyceride concentration higher than 443 mg/dL (5 mmol/L) were excluded (n=5).
Apolipoprotein E (ApoE) levels were determined in 2005 in one batch of plasma samples that were collected at age 85 years at study baseline and stored frozen. Plasma ApoE levels were determined using a human ApoE-specific sandwich enzyme-linked immunosorbent assay (ELISA) essentially as described (van Vlijmen et al., 1994). The detailed procedure is described in (Mooijaart et al., 2006).
Cytokine production capacity of the innate immune system
The cytokine production capacity of the innate immune system was assessed by stimulating ex vivo whole-blood samples with lipopolysaccharide (LPS) as described elsewhere (van der Lin- den et al., 1998). In short, all venous blood samples were drawn in the morning before 11 AM to exclude circadian variation, diluted 2-fold with RPMI-1640, and stimulated with Escherichia coli–derived LPS (10 ng/mL; Difco Laboratories, Detroit, MI). After 4 hours and after 24 hours of incubation at 378 C0 and 5 % CO2, supernatants were collected and stored at -808 C0 to mea- sure tumor necrosis factor-alpha (TNF-a), interleukin-1 beta (IL-1b), IL-6, IL-10, IL-12, IL-1 receptor antagonist (IL-1Ra), and interferon-gamma (IFN-c), respectively. Standard ELISA techniques were performed according to the manufacturer’s guidelines (Central Laboratory of the Blood Transfusion Service, Amsterdam, The Netherlands). Because of a possible distortion
by frailty (van den Biggelaar et al., 2004b), we restricted these analyses to those participants who survived for at least 2 years (n=463).
SNP selection
Four SNPs in the NR1H3 gene were selected using the HapMap database (http://www.hapmap.
org; version June 2005). Only validated SNPs were selected, and calculations on linkage disequi- librium (LD) and frequencies were performed using data from the European Centre d’Etude du Polymorphisme Humain (CEPH) population. As boundaries, the first SNP upstream of the ATG start site (LXRA5 untranslated region [UTR], rs11039149) and the first SNP downstream of the stop codon (LXRA3UTR, rs1449627) were selected. The expected D’ between these two SNPs was 1, indicating that the entire region is in strong LD. We additionally selected one SNP in exon 3 (LXRAex3, rs227923) and one in intron 6 (LXRAint6, rs712011), resulting in a set of four evenly spaced SNPs, separated by 5 kb.
Genotyping
The polymorphisms were genotyped using either an Assay-by-Design or an Assay-on-De- mand (Applied Biosystems, Nieuwerkerk aan den IJssel, The Netherlands), consisting of PCR primers and TaqMan Major Groove Binding (MGB) probes. For LXRA5UTR an Assay-by- Design was used with forward primer GAGCATCTGCAGGGTTCTCA, reverse primer GCCA- GTGAAGTGCTGTAATGGAA, one probe CCCCTGTAGCCCACC labeled with VIC, and one probe CCCTGTGGCCCACC labeled with FAM. For the LXRAex3 SNP an Assay-on-Demand was used with identification number C_15967384_10. For LXRAint6 an Assay-on-Demand was used with identification number C_1301060_20. For LXRA3UTR, an Assay-by-Design was used with forward primer CCTCACGTGCATGTGTAGCAT, reverse primer AGGTCTTTCAG- GTTGTGCCTTTT, one probe CCTTGGTTTTTCC labeled with VIC, and one probe CCTTG- GGTTTTCC labeled with FAM. Amplification reactions were performed at standard conditions except for the following modifications. A qPCR core kit was used (Eurogentec, Maastricht, The Netherlands) with half of the amount of primers and probes. Real-time PCR was performed on an ABI 7900 HT (Applied Biosystems), and genotypes were called using the Sequence Detection System 2.1 (Applied Biosystems). A random 10 % of all genotypes were performed in duplicate, and genotyping errors were < 2 % for all assays.
Statistical analysis
The program Haploview (Barrett et al., 2005) was used to estimate allele frequencies, test the consistency of genotype frequencies at each SNP locus with Hardy–Weinberg equilibrium, and estimate and plot pairwise LD between the SNPs examined. LD was estimated for all two-way comparisons of individual SNPs using two common measures: the r2 (the square of the stan- dardized correlation coefficient) and the Lewontin D’ (D’=D/Dmax if D > 0 or D’=D/Dmin if D < 0). Haplotypes and haplotype frequencies were estimated using the SNPHAP software (http://www-gene.cimr.cam.ac.uk/clayton/software). The posterior probabilities of pairs of haplotypes per subject as estimated by SNPHAP, were used as weights in the following analyses.
Continuous variables were normally distributed, except for plasma ApoE levels, triglycerides, CRP levels, and cytokines, which therefore were ln-transformed. All analyses were sex adjusted, using homozygosity for the most common haplotype as the referent group. Associations be- tween haplotypes and metabolic profile were analyzed using linear regression. Mortality risks and 95 % confidence intervals (CI) were calculated with the Cox proportional hazard model.
These analyses included all the estimated haplotypes in the model weighted for probability, ex- cept the reference haplotype. Clustered robust standard errors were calculated using individuals as clustering variable. These models assume an additive effect of the haplotypes. Haplotypes with low frequencies (< 5 %) fully participated in these analyses, but results on these haplotypes are not reported as their accuracy is low due to small numbers. The analyses were performed using STATA statistical software, version 9.0 (STATA Corp., College Station, TX).
Results
The baseline characteristics of the study populations are listed in Table 1. All participants were aged 85 years, and 67 % were female.
The position of the selected SNPs relative to the gene structure is shown in Figure 1a. The SNPs were in strong LD (D’ > 0.97) and constituted one haplotype block (Figure 1b) with seven haplotypes, of which the predicted frequencies are listed in Table 2. For the present analyses
Table 1. Baseline characteristics of the study population
Characteristic Value
Total number 563
Female (n, %) 375 (67 %)
Age (mean, SD)* 85.0 (-)
Lipid and lipoprotein plasma level
Total cholesterol, mean mmol/L (SD) 5.71 (1.13)
LDL cholesterol, mean mmol/L (SD) 3.68 (0.97)
HDL cholesterol, mean mmol/L (SD) 1.31 (0.40)
Triglycerides, median mmol/L (IQR) 1.34 (1.01–1.95)
CRP, median mg/L (IQR) 4.00 (1.00–8.00)
LPS-stimulated cytokines
IL-1b, median ng/mL (IQR) 3.50 (2.10–6.50)
IL-1RA, median ng/mL (IQR) 30.8 (28.3–46.0)
IL-6, median ng/mL (IQR) 60.7 (43.2–82.9)
IL-10, median pg/mL (IQR) 762 (487–1089)
IL-12, median ng/mL (IQR) 6.70 (4.30–10.20)
IFN-c, median ng/mL (IQR) 139 (43.0–448)
TNF-a, median ng/mL (IQR) 10.3 (7.40–13.3)
*All participants were enrolled within the month of their 85th birthday; SD=standard deviation; LDL=low-density lipoprotein;
HDL=high-density lipoprotein; IQR=interquartile range; CRP=C-reactive protein; LPS=lipopolysaccharide; IL=interleukin;
IFN=interferon; TNF=tumor necrosis factor; RA=receptor antagonist
we report the results of the four most common haplotypes (frequency > 5 %) that cumulatively account for > 97% of the haplotypes.
During a mean follow-up time of 4.9 years, 320 participants (57 %) had died. We compared the mortality risk associated with the various haplotypes, using the most common haplotype 1 as the reference category (Figure 2). The mortality risk was lower for haplotype 2 compared to haplotype 1 (hazard ratio [HR] = 0.78; 95 % CI, 0.64–0.95; p=0.015), whereas other haplotypes were not significantly associated with a higher or lower mortality risk. When assessing specific causes of death, the lower mortality risk that was associated with haplotype 2 was mainly caused by a lower mortality risk from infectious disease (HR=0.31, 95 % CI, 0.12–0.85; p=0.023) and from mortality in the category ‘‘other causes’’ (HR= 0.71, 95 % CI, 0.50–1.00; p=0.052).
Table 2. Haplotype structures and frequencies
SNP allele
Haplotype LXRa5UTR LXRaex3 LXRaint6 LXRa3UTR Frequency
1 A C T T 0.367
2 G C T T 0.273
3 A T C G 0.176
4 A C C G 0.164
5 A C T G 0.016
6 A T C T 0.004
7 A C C T 0.003
Minor alleles are depicted in bold. Minor allele frequencies of the four polymorphisms were: LXRa5UTR, 0.27 (G); LXRaex3, 0.18 (T); LXRaint6, 0.35 (C); LXRa3UTR, 0.36 (G). All genotype distributions were in Hardy–Weinberg equilibrium, SNP=single nucleotide polymorphism; LXR=liver X receptor; UTR=untranslated region; ex = exon; int=intron
Figure 1. Haplotype structure of the NR1H3 gene in the Leiden 85-plus Study population. (A) Relative position of the selected single nucleotide polymorphisms (SNPs) in the NR1H3 gene. Ver- tical bars: exons 1–9. Arrows: positions of SNPs.
(B) Visual representation of linkage disequilibri- um within the gene. Based on standard confidence interval criteria, all four SNPs constitute one hap- lotype block. Top triangles (pointing upwards, all black) indicate strong linkage disequilibrium of all four SNPs. Pairwise linkage disequilibrium is in- dicated by the numbers in the top triangles. Pair- wise R2 values are indicated by the numbers in the bottom triangles (pointing downwards), in which light gray triangles indicate low R2, and dark gray triangles indicate high R2.
The relationship between the four common haplotypes and variables in lipid metabolism is shown in Table 3. Haplotype 2 associated with significantly higher plasma ApoE levels (+0.48 mg/dL, p=0.018) and triglyceride levels (+0.098 mmol/dL, p=0.041) compared to haplotype 1.
Haplotype 4 also associated with higher plasma ApoE levels compared to haplotype 1, although the association was borderline statistically significant (+0.45 mg/dL, p=0.057), possibly due to the lower haplotype frequency.
To explore the association of NR1H3 haplotypes with innate immune function, we assessed cytokine production capacity by ex vivo whole-blood LPS-stimulated cytokine levels (Table 4).
We found no association of any haplotype with cytokine production capacity. Finally, to investi- gate the possibility that the LXRA regulates inflammation through alternative mechanisms, we associated the haplotypes with circulating CRP level, a plasma marker of systemic inflammation.
We found no association for any haplotype with circulating levels of CRP (Table 4).
Discussion
In C. elegans, the NHR DAF-12 has been shown to be one of the key components that modu- late lifespan in response to environmental cues. Based on protein sequence comparisons, we recently identified the LXRA as one of the human NHRs most similar to C. elegans DAF-12 (Mooijaart et al., 2005). Here we report that genetic variation in the gene coding for the LXRA (NR1H3) associates with human life span.
We found that a common haplotype of the NR1H3 gene associated with lifespan extension, predominantly attributable to decreased death from infectious disease. The LXRA is involved in various processes that contribute to infectious disease. The LXRA regulates specific processes that increase resistance to pathogens. For instance, LXRA regulates the expression of APa, a scavenging receptor that inhibits macrophage apoptosis and promotes the killing of the bacteria (Joseph et al., 2004). Although LXR agonists reduce inflammatory gene expression in models Figure 2. Dots represent hazard ratios calculated using a Cox proportional model adjusting for gender;
bars represent 95 % confidence intervals. Mean follow-up time was 4.9 years, in which time a total number of 320 participants (57 %) had died.
Table 3. Association of NR1H3 haplotypes with parameters of lipid metabolism
Haplotype 1Haplotype 2Haplotype 3Haplotype 4Plasma Component Mean(95 % CI) Difference(95 % CI)1p-value1Difference(95 % CI)1p-value1Difference(95 % CI)1p-value1
ApolipoproteinApoE (mg/dL)24.95 (4.52–5.42)0.48 (0.08-0.91)0.0180.30 (-0.13-0.79)0.1880.45 (-0.01-0.96)0.057LipidsTotal cholesterol, mmol/L5.93 (5.74–6.13)0.06 (-0.10-0.21)0.473-0.021 (-0.18-0.14)0.802-0.10 (-0.29-0.09)0.294LDL cholesterol, mmol/L3.84 (3.67–4.02)0.02 (-0.12- 0.15)0.806-0.072 (-0.21-0.07)0.311-0.10 (-0.27-0.06)0.226HDL cholesterol, mmol/L1.39 (1.32–1.46)-0.01 (-0.07-0.20)0.813-0.004 (-0.07-0.06)0.890-0.01 (-0.08-0.06)0.813Triglycerides, mmol/L21.38 (1.28–1.49)0.10 (0.00-0.20)0.0410.07 (-0.04-0.19)0.1950.02 (-0.08-0.14)0.662
1 Compared to Haplotype 1; 2 Geometric mean; Freq= frequency; ApoE=apolipoprotein E; LDL=low-density lipoprotein; HDL=high-density lipoprotein. Data represent sex-adjusted means and 95% confidence intervals (CI). All participants were aged 85 years
Table 4. Association of NR1H3 haplotypes with whole-blood lipopolysaccharide (LPS)-stimulated cytokine levels at baseline
Haplotype 1Haplotype 2Haplotype 3Haplotype 4Mean(95 % CI) Difference(95 % CI)1p-value1Difference(95 % CI)1p-value1Difference(95 % CI)1p-value1
Innate immunity2IL-1b, ng/mL3.30 (2.80-3.80)0.10 (-0.30-0.60)0.680-0.20 (-0.60-0.30)0.4400.40 (-0.10-1.00)0.141IL1-RA, ng/mL35.4 (32.4-38.6)-1.20 (-3.70-1.40)0.3400.10 (-2.80-3.20)0.972-0.20 (-3.30-3.30)0.859IL-6, ng/mL57.4 (51.8-63.4)-0.90 (-5.20-3.70)0.682-1.40 (-6.20-3.90)0.5920.60 (-4.30-6.10)0.794IL-10, pg/mL709 (623-807)5.00 (-64.0-82.0)0.888-6.00 (-85.0-83.0)0.892-19.0 (-101.0-75.0)0.680IL-12, ng/mL6.00 (5.20-6.90)0.20 (-0.40-1.00)0.494-0.60 (-1.30-0.01)0.0540.50 (-0.30-1.30)0.249IFN-c, ng/mL151 (108-209)10.0 (-27.0-59.0)0.617-35.0 (-66.0-8.0)0.10015.0 (-32.0-82.0)0.575TNF-a, ng/mL9.80 (9.00-10.7)0.40 (-1.10-0.30)0.240-0.80 (-1.60-0.10)0.0350.30 (-0.50-1.30)0.432Chronic inflammation2CRP, mg/L2.00 (1.50-2.70)0.20 (-0.30-0.80)0.4400.40 (-0.20-1.10)0.2240.20 (-0.40-0.90)0.599
1 Compared to Haplotype 1; 2 Geometrical means; Freq=frequency; IL=interleukin; RA=receptor antagonist; IFN=interferon; TNF=tumor necrosis factor; CRP=C-reactive protein. Data represent sex-adjusted means and 95% confidence intervals (CI). All participants were aged 85 years
of dermatitis and atherosclerosis (Joseph et al., 2003), LXRA-/- knockout mice are more suscep- tible to infection with Listeria monocytogenes (Joseph et al., 2004).
In the search for an intermediate phenotype, we associated genetic variation in the NR1H3 gene with mortality-related markers in lipid metabolism and immunity. We selected these mark- ers because these phenotypes are known to associate with mortality and a functional relation- ship with the LXRA protein was plausible. Cholesterol metabolism is related to various causes of death (Weverling-Rijnsburger et al., 1997), and the LXRA is involved in various components of lipid metabolism, such as reverse cholesterol transport, cholesterol excretion, and fatty acid synthesis (Lu et al., 2001). We observed an association of haplotype 2 of the NR1H3 gene with increased plasma ApoE and triglyceride levels. ApoE is a component of triglyceride-rich lipoproteins, such as very low- density lipoprotein (VLDL), which may explain why haplotype 2 associates with plasma levels of both ApoE and triglycerides. Furthermore, LXRA agonists have been suggested for therapeutic use against cardiovascular disease, but a serious side effect of the use of LXR agonists as therapeutic agents is the concomitant increase in liver and serum triglycerides (Grefhorst et al., 2002). These effects are caused by a strong induction of lipogenic genes in the liver and an increased VLDL production (Grefhorst et al., 2002). In line with these animal data, we found that haplotype 2 also associated with higher ApoE levels. Thus, the as- sociations with increased levels of triglycerides and ApoE presented here suggest that haplotype 2 associates with increased LXRA activity in the liver. It is interesting that triglyceride-rich lipo- proteins (of which ApoE is a component) act as agents of the innate immune system (Barcia and Harris, 2005), for example, by binding and neutralizing bacterial components. ApoE redirects lipopolysaccharides (bacterial components) in the liver from Kupffer cells to hepatocytes and protects against endotoxemia in rats (Rensen et al., 1997). Recently, it was discovered that ApoE is also involved in lipid antigen presentation (van den et al., 2005) and that high plasma ApoE levels associate with increased systemic inflammation (Mooijaart et al., 2006).
LPS-stimulated cytokine production levels are highly heritable (de Craen et al., 2005), and cytokine production profiles associate with patterns of old age mortality (van den Biggelaar et al., 2004a). However, genetic variation in the genes coding for the cytokines has so far been insufficient to explain the heritable component (Haukim et al., 2002). In the present study, varia- tion in the NR1H3 gene did not associate with ex vivo LPS-stimulated whole-blood cytokine levels or with circulating CRP levels. Others have demonstrated an association of the LXRA with inflammation in macrophage and monocyte cell cultures (Joseph et al., 2003; Landis et al., 2002). However, inflammatory cytokines and other serum mediators were not different between LXRA and LXRB knockout mice and wild types (Joseph et al., 2004). We interpret that NR1H3 may not be a major determinant of cytokine production capacity in blood upon stimulation by LPS. This interpretation does not, however, exclude the possibility that in other cell types NR1H3 haplotypes may influence the local production of cytokines.
We did not observe a beneficial effect of haplotype 2 on death from cardiovascular causes.
In mouse models, LXR agonists reduce the formation of atherosclerotic lesions (Joseph et al., 2002) whereas macrophage-specific LXRA knockout aggravates atherosclerosis development (Tangirala et al., 2002). To date, the function of the LXRA in lipid and cholesterol metabolism has been studied in mouse models and in human cultured cell lines, mostly macrophages. How- ever, caution must be taken in extrapolating these results based on cultured cells and mouse models of atherosclerosis to humans (Repa and Mangelsdorf, 2002). Whereas macrophage LXR
has been shown to be antiatherogenic (Levin et al., 2005; Tangirala et al., 2002), this beneficial effect on cardiovascular disease may be balanced by proatherogenic effect of liver LXR activa- tion.
Genetic variation in both C. elegans daf-12 and the human NR1H3 gene associates with dif- ferences in life span, suggesting that the two genes may, at least to some extent, be functionally conserved. Other evolutionarily conserved pathways have previously been implicated in life- span regulation. For example, it was first discovered that the C. elegans daf-2 mutant was long- lived (Kenyon et al., 1993). Later it was discovered that the daf-2 gene shows homology to the mammalian genes encoding the insulin receptor (IR) and insulin-like growth factor 1 receptor (IGF-1R) (Kimura et al., 1997), which are conserved throughout evolution. Extended life span was then also demonstrated in IR mutants in Drosophila melanogaster (Tatar et al., 2001) and in IR and IGF-1R mutants in mice (Bluher et al., 2003; Holzenberger et al., 2003). Recently we showed that reduced insulin signaling in humans also associates with longevity (van Heemst et al., 2005).
These observations suggest that the approach of studying evolutionarily conserved pathways is fruitful in identifying genes that regulate human life span.
Very recently, several articles report on the biological function of daf-12 in C. elegans. Two hormones were identified that function as DAF-12 ligands (Motola et al., 2006), and the bio- synthetic pathway of production of these ligands was described in more detail (Rottiers et al., 2006). Furthermore, cholestenoic acid was found to rescue the worm from dauer diapause in a DAF-12–dependent manner (Held et al., 2006). These studies provide important new hints to investigate the functional conservation of life-span regulation throughout evolution and the biological function of the human LXRs or other NHRs.
A limitation of our study is that it does not include analyses of the gene encoding the LXRB (NR1H2). The LXRA and LXRB are highly similar proteins, as their amino acid sequences are very alike and the proteins have similar functions in lipid metabolism and inflammation. It could therefore be hypothesized that a loss of function of one of the genes will be compensated for by the function of the other receptor and will therefore not have dramatic effects. The LXRA and LXRB are encoded by different genes and have different expression patterns. Whereas the LXRA is expressed only in a limited number of tissues, the LXRB is expressed ubiquitously.
Indeed, in activated macrophages the inhibiting effect of LXR ligands on cytokine expression is completely abrogated in double knockout macrophages (nr1h3-/- nr1h2-/-). However, there was also a partial reduction of this effect in nr1h3-/- macrophages (Joseph et al., 2003). Furthermore, whereas LXRA-/- knockout increases susceptibility to bacterial infection, additional knockout of LXRB does not increase this susceptibility (Jonsson et al., 2003). These observations sug- gest that genetic variation in LXRA may have functional significance independent of LXRB. A second limitation is that it is unclear which genetic variant within the haplotype is responsible for the observed functional variation. The SNP that tags haplotype 2 is located in the 5` UTR of the gene. It is, however, unknown whether this SNP itself has functional significance—for instance, by affecting promoter function or interaction domains. Alternatively, the 5` UTR SNP may be in LD with an SNP elsewhere in the genome, for instance in the coding region of the gene, which leads to an alteration with functional significance.
The strong point of our study is that we selected genetic variants tagging all common hap- lotypes of the NR1H3 gene and associated them with a range of variables in inflammation and lipid metabolism. Furthermore, the prospective nature of the mortality analyses and the rela-
tively large number of deaths yield a powerful tool to study mortality risk.
As this is the first report on the effect of common genetic variants in the human NR1H3 gene on human life span, the results of our observational study need further replication. Furthermore, more research is warranted to confirm which specific genetic variation on the haplotypes actu- ally changes the function of the gene or protein.
Acknowledgments
This work was funded by an IOP (Innovative Oriented Research) grant (IGE01014) from the Dutch Ministry of Economic Affairs. We thank Dennis Kremer for technical assistance.
References
Antebi A, Culotti JG, Hedgecock EM (1998). daf-12 regulates developmental age and the dauer alterna- tive in Caenorhabditis elegans. Development 125: 1191-1205.
Antebi A, Yeh WH, Tait D, Hedgecock EM, Riddle DL (2000). daf-12 encodes a nuclear receptor that regulates the dauer diapause and developmental age in C. elegans. Genes Dev 14: 1512-1527.
Barcia AM, Harris HW (2005). Triglyceride-rich lipoproteins as agents of innate immunity. Clin Infect Dis 41 Suppl 7: S498-S503.
Barrett JC, Fry B, Maller J, Daly MJ (2005). Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 21: 263-265.
Bluher M, Kahn BB, Kahn CR (2003). Extended longevity in mice lacking the insulin receptor in adi- pose tissue. Science 299: 572-574.
de Craen AJ, Posthuma D, Remarque EJ, van den Biggelaar AH, Westendorp RG, Boomsma DI (2005).
Heritability estimates of innate immunity: an extended twin study. Genes Immun 6: 167-170.
der Wiel AB, van Exel E, de Craen AJ, Gussekloo J, Lagaay AM, Knook DL et al (2002). A high response is not essential to prevent selection bias: results from the Leiden 85-plus study. J Clin Epidemiol 55: 1119-1125.
Fisher AL, Lithgow GJ (2006). The nuclear hormone receptor DAF-12 has opposing effects on Caenorhabditis elegans lifespan and regulates genes repressed in multiple long-lived worms. Aging Cell 5: 127-138.
Grefhorst A, Elzinga BM, Voshol PJ, Plosch T, Kok T, Bloks VW et al (2002). Stimulation of lipogen- esis by pharmacological activation of the liver X receptor leads to production of large, triglyceride- rich very low density lipoprotein particles. J Biol Chem 277: 34182-34190.
Haukim N, Bidwell JL, Smith AJ, Keen LJ, Gallagher G, Kimberly R et al (2002). Cytokine gene poly- morphism in human disease: on-line databases, supplement 2. Genes Immun 3: 313-330.
Held JM, White MP, Fisher AL, Gibson BW, Lithgow GJ, Gill MS (2006). DAF-12-dependent rescue of dauer formation in Caenorhabditis elegans by (25S)-cholestenoic acid. Aging Cell 5: 283-291.
Holzenberger M, Dupont J, Ducos B, Leneuve P, Geloen A, Even PC et al (2003). IGF-1 receptor regu- lates lifespan and resistance to oxidative stress in mice. Nature 421: 182-187.
Janowski BA, Grogan MJ, Jones SA, Wisely GB, Kliewer SA, Corey EJ et al (1999). Structural require- ments of ligands for the oxysterol liver X receptors LXRalpha and LXRbeta. Proc Natl Acad Sci U S A 96: 266-271.
Jonsson EG, Cichon S, Gustavsson JP, Grunhage F, Forslund K, Mattila-Evenden M et al (2003). As- sociation between a promoter dopamine D2 receptor gene variant and the personality trait detach- ment. Biol Psychiatry 53: 577-584.
Joseph SB, Bradley MN, Castrillo A, Bruhn KW, Mak PA, Pei L et al (2004). LXR-dependent gene ex- pression is important for macrophage survival and the innate immune response. Cell 119: 299-309.
Joseph SB, Castrillo A, Laffitte BA, Mangelsdorf DJ, Tontonoz P (2003). Reciprocal regulation of in- flammation and lipid metabolism by liver X receptors. Nat Med 9: 213-219.
Joseph SB, McKilligin E, Pei L, Watson MA, Collins AR, Laffitte BA et al (2002). Synthetic LXR li- gand inhibits the development of atherosclerosis in mice. Proc Natl Acad Sci U S A 99: 7604-7609.
Kenyon C, Chang J, Gensch E, Rudner A, Tabtiang R (1993). A C. elegans mutant that lives twice as long as wild type. Nature 366: 461-464.
Kimura KD, Tissenbaum HA, Liu Y, Ruvkun G (1997). daf-2, an insulin receptor-like gene that regu-
lates longevity and diapause in Caenorhabditis elegans. Science 277: 942-946.
Klass M, Hirsh D (1976). Non-ageing developmental variant of Caenorhabditis elegans. Nature 260:
523-525.
Landis MS, Patel HV, Capone JP (2002). Oxysterol activators of liver X receptor and 9-cis-retinoic acid promote sequential steps in the synthesis and secretion of tumor necrosis factor-alpha from human monocytes. J Biol Chem 277: 4713-4721.
Levin N, Bischoff ED, Daige CL, Thomas D, Vu CT, Heyman RA et al (2005). Macrophage liver X receptor is required for antiatherogenic activity of LXR agonists. Arterioscler Thromb Vasc Biol 25: 135-142.
Lu TT, Repa JJ, Mangelsdorf DJ (2001). Orphan nuclear receptors as eLiXiRs and FiXeRs of sterol metabolism. J Biol Chem 276: 37735-37738.
Mooijaart SP, Berbee JF, van Heemst D, Havekes LM, de Craen AJ, Slagboom PE et al (2006). ApoE plasma levels and risk of cardiovascular mortality in old age. PLoS Med 3: e176.
Mooijaart SP, Brandt BW, Baldal EA, Pijpe J, Kuningas M, Beekman M et al (2005). C. elegans DAF- 12, Nuclear Hormone Receptors and human longevity and disease at old age. Ageing Res Rev 4:
351-371.
Motola DL, Cummins CL, Rottiers V, Sharma KK, Li T, Li Y et al (2006). Identification of ligands for DAF-12 that govern dauer formation and reproduction in C. elegans. Cell 124: 1209-1223.
Peet DJ, Turley SD, Ma W, Janowski BA, Lobaccaro JM, Hammer RE et al (1998). Cholesterol and bile acid metabolism are impaired in mice lacking the nuclear oxysterol receptor LXR alpha. Cell 93:
693-704.
Rensen PC, Oosten M, Bilt E, Eck M, Kuiper J, Berkel TJ (1997). Human recombinant apolipoprotein E redirects lipopolysaccharide from Kupffer cells to liver parenchymal cells in rats In vivo. J Clin Invest 99: 2438-2445.
Repa JJ, Mangelsdorf DJ (2002). The liver X receptor gene team: potential new players in atheroscle- rosis. Nat Med 8: 1243-1248.
Repa JJ, Turley SD, Lobaccaro JA, Medina J, Li L, Lustig K et al (2000). Regulation of absorption and ABC1-mediated efflux of cholesterol by RXR heterodimers. Science 289: 1524-1529.
Rottiers V, Motola DL, Gerisch B, Cummins CL, Nishiwaki K, Mangelsdorf DJ et al (2006). Hormonal control of C. elegans dauer formation and life span by a Rieske-like oxygenase. Dev Cell 10: 473- 482.
Schoenmaker M, de Craen AJ, de Meijer PH, Beekman M, Blauw GJ, Slagboom PE et al (2006). Evi- dence of genetic enrichment for exceptional survival using a family approach: the Leiden Longevity Study. Eur J Hum Genet 14: 79-84.
Tangirala RK, Bischoff ED, Joseph SB, Wagner BL, Walczak R, Laffitte BA et al (2002). Identifica- tion of macrophage liver X receptors as inhibitors of atherosclerosis. Proc Natl Acad Sci U S A 99:
11896-11901.
Tatar M, Kopelman A, Epstein D, Tu MP, Yin CM, Garofalo RS (2001). A mutant Drosophila insulin receptor homolog that extends life-span and impairs neuroendocrine function. Science 292: 107- 110.
van den Biggelaar AH, de Craen AJ, Gussekloo J, Huizinga TW, Heijmans BT, Frolich M et al (2004a).
Inflammation underlying cardiovascular mortality is a late consequence of evolutionary program- ming. FASEB J 18: 1022-1024.
van den Biggelaar AH, Huizinga TW, de Craen AJ, Gussekloo J, Heijmans BT, Frolich M et al (2004b).
Impaired innate immunity predicts frailty in old age. The Leiden 85-plus study. Exp Gerontol 39:
1407-1414.
van den EP, Garg S, Leon L, Brigl M, Leadbetter EA, Gumperz JE et al (2005). Apolipoprotein- mediated pathways of lipid antigen presentation. Nature 437: 906-910.
van der Linden MW, Huizinga TW, Stoeken DJ, Sturk A, Westendorp RG (1998). Determination of tumour necrosis factor-alpha and interleukin-10 production in a whole blood stimulation system:
assessment of laboratory error and individual variation. J Immunol Methods 218: 63-71.
van Heemst D, Beekman M, Mooijaart SP, Heijmans BT, Brandt BW, Zwaan BJ et al (2005). Reduced insulin/IGF-1 signalling and human longevity. Aging Cell 4: 79-85.
van Vlijmen BJ, van den Maagdenberg AM, Gijbels MJ, van der BH, HogenEsch H, Frants RR et al (1994). Diet-induced hyperlipoproteinemia and atherosclerosis in apolipoprotein E3-Leiden trans- genic mice. J Clin Invest 93: 1403-1410.
vB HJ, Iachine I, Skytthe A, Vaupel JW, McGue M, Koskenvuo M et al (2006). Genetic influence on human lifespan and longevity. Hum Genet 119: 312-321.
Weverling-Rijnsburger AW, Blauw GJ, Lagaay AM, Knook DL, Meinders AE, Westendorp RG (1997).
Total cholesterol and risk of mortality in the oldest old. Lancet 350: 1119-1123.
